Protein Sequence Clusters

3 posts / 0 new
Last post
BioFerG
BioFerG's picture
Protein Sequence Clusters

 HI, everyone.
Sorry if my question is a little silly, but i have the following problem.
I aligned some sequences from a family gene according to an specific domain of the protein. I`m not include the entire protein because the others domains are less conserved. In my work, I did annotations using Blast and Interproscan to confirm, besides confirm important motifs using MEME/MAST.  I clustered the phylogenetic tree in order to observed relationships in this domain and cluster are well supported with bootstrapping values over 80% in the informative branches to me and in a many few of them over 68, that it was the minimum value obtained. When i look inside the clusters formed, the cluster show an identity values from 29 %. According to my work I think that the clustering its rigth, and I think despite of this values, if I can obtained the similarity values, i can support the cluster better, because i can say that this clusters are reflecting functional relationships, but i don't know how to obtained this values to report it. Can you help me? Please? the question is How can i obtained the similarity values among sequences in a multiple sequence alignment of protein sequences?
 
Thank you.
   

ryan_m
ryan_m's picture
Hi there.

Hi there.
If you know the % identity between the sequences, what kind of similarity values are you after?  Are you looking for another way to measure the homology of these sequences?
 
Ryan

BioFerG
BioFerG's picture
 

 
Absolutely, I am just looking for similarity values, correct me if I wrong. The identity value just have into account just the exactly aminoacid match, and similarity value depending of the substitution matrix used in the alignment have into account synonymous aminoacid match. I want this value for my alignment. Or do you think that its a better way to do it?, I am trying to measure the homology as you say with a more informative value than identity, if you have an advice about that I really appreciate it. Thank you so much, and my apologize if this question is too silly. I am new in this issue.
 
ryan_m wrote:

Hi there.
If you know the % identity between the sequences, what kind of similarity values are you after?  Are you looking for another way to measure the homology of these sequences?
 
Ryan